Community detection in complex networks

Abstract

Over the last decade, complex networks have emerged to be a promising research field. It enables us to extract and comprehend a variety of real systems, ranging from biology, technology, and sociology. For example, we can explore social network analysis, which involves thousands of relationships between people, or we can look at the protein interaction network to gain a deeper understanding of fundamental cellular biochemistry and physiology. To take advantage of the vast benefits that these systems bring, we can predict and perhaps control them by understanding their mathematical descriptions, but the effort is difficult due to the complexity of these systems and the significant differences between them. Even though these complex systems varied, several researchers discovered that many network architectures were quite similar because they were all built using the same organizing principle—they formed into densely linked communities. Community detection methodology, which can discover and cluster groups of nodes, has emerged as one of the fundamental approaches to address this issue. The problem of community detection in networks caught the attention of applied mathematicians and physicists all around the world, and several creative solutions were developed in an attempt to tackle it. The background, algorithms, and techniques for identifying communities in complex networks will all be thoroughly explored in this thesis report. Additionally, an experiment of how the community detection technique can improve the performance of the machine learning model in real-world scenarios will also be provided, along with in-depth discussions and investigations.